Exploring the limitations of the Markowitz portfolio optimisation model

Visiting Professor Paul Bilokon of Imperial College discusses the limitations of Harry Markowitz's 1952 Portfolio Optimization Model. He delves into issues such as non-stationarity in financial time series and proposes modern AI-based solutions like reinforcement learning and multi-armed bandits to address these challenges. The conversation also explores the impact of AI on the quantitative finance workforce, the critical need for explainable AI, and strategies for synergistic human-AI collaboration.