Portfolio optimization with AI: Evaluating Performance Beyond Traditional Techniques

Abstract

This thesis study the use of artificial intelligence (AI) in portfolio optimization, by evaluating and comparing AI with traditional methods such as Modern Portfolio Theory and Capital Asset Pricing Model. Advanced machine learning techniques such as deep learning, reinforcement learning and natural language processing which comprise AI bring several possibilities to ad- dress this weakness of traditional models in dynamic and volatile financial environments. This paper evaluates the use of AI in enhancing risk-adjusted returns and market agility before de- fining its challenges such as computational burden and interpretability. Our key observations describe situations where AI approaches outperform traditional methods, offering insights into possible future uses in asset management and field trends.