Exploring the sustainable fixed income market: A clustering-based approach with evidence from Italy and France

In the context of the global climate change and sustainable development, Green, Social, Sustainability, Sustainability-Linked, and Transition bonds (GSS+) have emerged as crucial instruments for financing projects with environmental and social objectives. This paper proposes a methodological framework to analyze GSS+ bonds by considering their financial and non-financial characteristics and their alignment with the United Nations Sustainable Development Goals (SDGs). It introduces a reproducible and scalable clustering approach to study this segment of the market. We perform an initial international exploratory analysis using a global dataset of more than 50,000 bond issuances and then a detailed case study focused on Italy and France. The Partitioning Around Medoids (PAM) algorithm is applied for clustering, and bond dissimilarities are calculated using the Gower distance, which is appropriate for mixed data types. The Rand index is used to assess the robustness of the clustering structure. The findings show different national characteristics: Italy’s market is smaller and more concentrated, with a corporate focus on infrastructure and energy, while France presents a wider issuance landscape with a more interconnected use of proceeds and SDG structure.