Modelling extreme risks for green bond and clean energy.

Abstract

Value at Risk and Expected Shortfall are two traditional tools used to measure extreme risk in financial markets. However, there is little research on measuring extreme risk in emerging markets such as green bonds and clean energy. This paper uses both semi-parametric models with simultaneous excitation functions and traditional models to estimate extreme risk in SP500 Green Bond (GB) and Global Clean Energy (GCE), selecting Expected Shortfall (ES) and Value at Risk (VaR) as the indices of extreme risk. Then, the paper uses a breakpoint scan of the predictions of the different types of models. The results find that the green bond market was relatively stable while the global clean energy market experienced sharp fluctuations after the COVID-19 pandemic outbreak. Representative models in GCE have at least two breakpoints, but those for GB have no breakpoints. The GARCH model with normal innovations performs the best among all target models, and the GARCH-FZ model fits the best among all semi-parametric candidates. Our research could help governments, companies, and investors with risk management and improve model accuracy and mechanisms for measuring extreme risks.