Green finance and portfolio allocation: diversification benefits of green bonds and probabilistic modelling of climate risk

We propose an analysis of green bonds and their relationship with financial and climate risk. Two global green bond indices, Solactive and Bloomberg Barclays MSCI, are investigated, in terms of diversification benefits and co-movement with multiple market indices, particularly those of sectors affected by the Covid-19 pandemic. The Solactive Green Bond Index consistently outperforms during the last seven years, including the pandemic. The impact of green-bond indices on otherwise traditional portfolios is also analysed, with portfolios that include green bonds always proving preferable in terms of risk. As for returns, overall, non-green portfolios perform better in times of market growth, while green portfolios achieve positive or less negative returns during downturns. Next, climate risk in the corporate and government bond market is studied, focusing on Eurozone green and non-green bonds. We evaluate the impact on spreads of multiple potential risk drivers, selected in line with the ECB climate stress tests and the extant literature, including EU carbon allowances as a proxy of transition risk. We find that green and non-green bonds of the same issuer can differ in their exposure to the physical risk variables. Green bonds can be equally, less exposed than their non-green counterparts, or even provide protection against physical risk. Carbon allowances have no consistent impact on bond spreads. Based on these results, we apply a credit risk model incorporating climate risk. The study concludes by exploring tail dependence, confirming the above findings for corporate bonds. For government bonds, it provides additional insight: green bonds from different countries exhibit divergent market treatment. Dutch and French green spreads closely mirror traditional counterparts, while German, Italian, and Spanish green bonds display less co-movement in tail events. Various explanations are suggested, including a "scarcity effect" linking the tail behaviour of green bonds to their scarcity relative to total government debt.