This is a repo for the paper: "A Machine Learning Approach to Systemic Risk: Combining CoVaR and Hierarchical Risk Parity"

Abstract: This study presents an innovative framework for systemic risk analysis by integrating Hierarchical Risk Parity (HRP) with the traditional Conditional Value-at-Risk (CoVaR) methodology. Leveraging daily stock price data from the STOXX 600 index over the past decade, the research applies machine learning techniques to compute CoVaR measures and uncover systemic risk drivers. The findings illustrate that CoVaR-based methods often concentrate risk within highly correlated company clusters, exacerbating systemic vulnerabilities. In contrast, HRP achieves a more balanced risk allocation, mitigating such vulnerabilities. The study offers critical insights into risk distribution, identifying interconnected clusters and outlier entities with unique risk profiles.