Abstract/Summary

This thesis explores financial risk measurement, specifically addressing market risk, climate transition risk, and credit risk, which are organized into three main chapters. The first contribution is that it proposes to estimate Value-at-Risk (VaR) and Expected Shortfall (ES) at extreme levels using information from a common level. We employ time series cross-validation to optimize the forecasting performance. Our simulation study reveals that the proposed novel models outperform the original Generalized Autoregressive Score model of Patton et al. (2019) according to various backtests. Empirical evidence based on the return data of four oil futures also shows the superior performance of the proposed models. Notably, our models’ performance is most prominent during the COVID-19 period. The second contribution is that it provides a framework to measure the effects of climate transition risk factors, proxied by the environmental pillar of ESG scores, on corporate downside risk. Analyzing the stock returns and climate risk factors relationship, a notable negative correlation in lower quantiles is revealed. A new risk measure for climate transition risk factors is also proposed, with empirical findings indicating sector-based variations in sensitivity to these risks. Specifically, the Health Care sector is the least efficient in reducing climate risk, while the Energy sector benefits the most from improvements in the firms’ environmental scores. The third contribution is a study examining how corporate environmental performance influences credit ratings, with a trans-Atlantic study encompassing firms from the United States (US) and the European Union (EU). We find that corporate environmental performance positively affects the firms’ credit ratings. Interestingly, our findings reveal a linear relationship in the US and a nonlinear one in the EU. These findings highlight the implications of environmental performance. They provide vital insights for firms aiming to improve their credit rating via sustainability initiatives, while considering regional disparities.