Hirschey, Nicholas H.Teodoro, Miguel Gonçalves2025-02-262025-02-262024-01-232024-01-23http://hdl.handle.net/10362/179845In this directed research, we look to capitalize on a tree gradient boosting model for stock volatility prediction. Through an extensive literature review we build on past research focusing on a wide range of volatility drivers and integrate them as model features. We propose different rebalancing techniques to the market portfolio according to our volatility predictions and assess their viability. Finally, we arrive at an approach that offers a robust framework for equity volatility forecasting and propose portfolio constructions that can further advance the current understanding on the use of volatility for fund managers.engVolatility timingSupervised learningXgboostPortfolio rebalancingmaster thesis203866134