Humans continue to hold an advantage over artificial intelligence in the realm of financial market profitability, according to recent research that challenges earlier findings on this topic.
AI has firmly established itself in the financial services industry over the past decade, impacting various areas from cybersecurity to customer service chatbots. However, despite the vast amounts of data generated in financial markets, machine learning and AI technologies have yet to take a significant role in investment decision-making.
In their comprehensive study, Barbara Jacquelyn Sahakian, Fabio Cuzzolin, and Wojtek Buczynski examined 27 peer-reviewed articles published between 2000 and 2018, focusing on stock market forecasting experiments utilizing machine-learning techniques. The researchers identified a tendency for selective reporting in these studies; many experiments featured multiple iterations of investment models, with the authors predominantly showcasing only the most successful outcomes as their primary findings.
The researchers noted, “This method is impractical for real-world investment management, where strategies can only be executed once, resulting in definitive profits or losses—there is no option to reverse outcomes.”
Additionally, challenges such as the opaque nature of AI algorithms and insufficient attention to regulatory considerations may contribute to the slow adoption of these technologies in financial markets. In fact, a review of AI-driven funds with publicly available performance data indicated that they typically underperformed compared to market benchmarks.
The researchers concluded that, at present, there is a compelling case for the superiority of human analysts and fund managers. Despite their flaws, the evidence strongly indicates that human decision-makers currently outpace AI in effectiveness.