Introduction

Novel applications of machine learning are driving significant advancements in the field of empirical asset pricing. This PhD project, supervised by Dr. Simon Rottke and Dr. Florian Peters, contributes to this area by investigating return predictability and the implications of predictive uncertainty for quantitative investors’ decision making. Applications of predictive models to the cross-section of established risk factors are examined. The research also explores novel signal generation techniques to enhance the understanding of market dynamics, including phenomena such as momentum and investor behavioural biases like under- and over-reaction. By addressing these questions, the project provides fresh insights into model robustness, practical applications, and the transformative role of machine learning in finance.