Analyzing the green bond index: A novel quantile-based high-dimensional approach

Abstract

The development of green bond markets is important for advancing energy efficiency, supporting renewable energy, encouraging sustainable investments, and safeguarding the environment. However, the inherent complexity and uncertainty of these markets pose significant challenges for both investors and researchers. In this study, we focus on analyzing the S&P Green Bond Index, a leading benchmark for monitoring the global green bond market. We introduce a new high-dimensional statistical method, the Quantile Group Adaptive Lasso, designed to accurately predict the returns of this index. Our empirical results demonstrate that this model surpasses several established forecasting techniques in both accuracy and stability. Furthermore, our analysis of economic significance highlights the critical influence of traditional energy-related predictors from G7 and BRICS countries on the global green bond markets. We also find that monetary policies and macroeconomic factors, such as M2 money supply, CPI, and government bond yields, play vital roles. Additionally, the robustness of our proposed method is confirmed. Overall, our study provides a powerful tool that not only significantly enhances forecasting performance but also deepens the understanding of the interplay between trends in green bond markets and information from energy sectors and broader economic conditions.