Uncertainty, Volatility Spillovers, and Sustainability in Financial Markets

This thesis investigates how stock market risk and volatility evolve in a changing global financial environment. Integrating insights from volatility forecasting, systemic risk transmission, and environmental sustainability, it applies advanced econometric and machine-learning methods to study how different forms of uncertainty are forecasted, transmitted, and priced in international markets. Chapter2 examines whether country-specific economic policy uncertainty (EPU) improves forecasts of realised volatility in the United States, United Kingdom, Japan, the Euro Area, and Hong Kong (2001–2022) within an HAR framework. Across markets, implied volatility is the strongest short-horizon predictor, while EPU provides additional forecasting value at longer horizons, supporting volatility timing and risk control. By contrast, the incremental contribution of EPU is limited in Hong Kong, consistent with local policy signals playing a less dominant role in volatility dynamics. Chapter3 analyses volatility spillovers between the G7 and major emerging markets (Brazil, India, China) from 2010 to 2022, with emphasis on systemic episodes such as COVID-19. Using connectedness measures in the time, quantile, frequency, and time-varying VAR domains, it documents persistently high spillovers that rise sharply in crises, event-driven shifts in net transmission roles, and long-horizon persistence. Regression evidence indicates that global risk sentiment (VIX) and economic policy uncertainty (EPU) are robust macro-financial indicators associated with fluctuations in aggregate connectedness, supporting surveillance and stress-scenario design. Chapter4 studies whether firms’ environmental sustainability performance predicts US aggregate market risk, focusing on the variance risk premium (VRP). Using 14 Sustainalytics indicators for S&P 500 firms, it constructs a market-level environmental index and shows out-of-sample predictive power for VRP, with emissions disclosure, carbon intensity (and its trend), and green procurement policies among the most informative predictors. Notably, stronger carbon performance is associated with higher VRP, consistent with transition-related uncertainty. Overall, the thesis clarifies how policy, systemic, and environmental uncertainties jointly shape modern financial risk and informs monitoring, portfolio management, and policy-oriented assessment.