Clustering of Green, Social, Sustainability, and Sustainability-Linked Bonds (2014–2023)

Abstract

Green, social, sustainability, and sustainability-linked bonds have become relevant mechanisms for channeling capital toward projects with environmental and
social benefits. However, much of the empirical literature has focused on market
growth, issuance credibility, and pricing, while less evidence is available on the
structural heterogeneity of the global market. The aim of this study is to analyze
the sustainable finance ecosystem by classifying green, social, sustainability, and
sustainability-linked bonds issued between 2014 and 2023 through clustering techniques. The study starts from a public database of 2,546 records and builds an analytical sample of 2,340 observations, for which the year variable is derived from the
Latest External Review Report field. Based on this sample, a descriptive analysis and
an unsupervised segmentation using K-means were performed on transformed and
standardized quantitative variables. The results show the predominance of green
bonds in both frequency and aggregate amount, a marked geographic and sectoral
concentration, and three distinct segments differentiated by issuance scale and recurrence: small-scale and low-recurrence issuances, large-scale and high-recurrence
issuances, and intermediate-scale issuances with moderate recurrence. This study
provides a structural and reproducible reading of the sustainable debt market that
may serve as a basis for future research on spreads, external review credibility, and
financial performance.