Essays on portfolio optimization and volatility forecasting via high-frequency financial data

This thesis advances the literature on portfolio optimization and volatility forecasting using high-frequency financial data. Chapter Two examines the utility of well-known realized measures in portfolio optimization. Chapter Three focuses on the dynamics of drift and its role in volatility forecasting, particularly for highly volatile assets like cryptocurrencies. Chapter Four explores the use of two key non-normal features of asset price variation—jumps and drifts—for improved volatility forecasting. By addressing these critical aspects, the thesis contributes to a deeper understanding of financial dynamics and enhances methodologies for asset management and risk assessment.<p></p>